|CHAPTER 5 MEDICAL USE OF CANNABIS AND CANNABINOIDS:
REVIEW OF THE EVIDENCE
5.26 Besides MS, the other main indication claimed
for cannabis-based medicines is the control of pain (analgesia). The BMA report says,
"The prescription of nabilone, THC and other cannabinoids...should be permitted for
patients with intractable pain", especially in terminal illness.
5.27 Professor Wall told us that there is clear
evidence of analgesic effects of cannabis and cannabinoids from animal experiments. Some
of the results suggest that pain which originates from damaged nerves might respond to
cannabinoids; this could be of medical value as this type of pain does not respond well to
treatment with morphine and related narcotics (Q 99). An example of such pain is
phantom limb pain following amputation (Q 100). As many as 30 per cent of amputees
suffer from this distressing condition, for which there is currently no satisfactory
treatment. Dr Colin Stewart, who works in the field of major limb amputation in
Dundee, reports anecdotal evidence that cannabis can relieve this pain; he recommends that
trials of cannabis be undertaken in such patients (p 304).
5.28 Dr David Lambert, of the University of
Leicester, confirmed that there is evidence for analgesic actions of cannabinoids acting
on both the spinal cord and higher brain centres. He and Dr Notcutt suggested that
one way of dissociating the painrelieving actions of cannabinoids from their
psychoactive effects might be to deliver the cannabinoid locally to the spinal cord via
the cerebrospinal fluid, as has been done with opiate analgesics (QQ 440-6).
5.29 Dr Anita Holdcroft of Hammersmith Hospital, a
contributing author to the BMA report, has reported the results of a placebo-controlled
trial of cannabis in a patient with severe chronic pain of gastrointestinal origin
(diagnosed as familial Mediterranean fever).
Treatment was with capsules of cannabis oil, standardised for THC content. The patient's
demand for morphine was substantially lower during treatment with cannabis than during a
period of placebo treatment (p 224).
5.30 In short, there is scientific evidence that
cannabinoids possess painrelieving properties, and some clinical evidence to support
their medical use in this indication. Many of our witnesses consider that high priority
should be given to further research in this area.
5.31 There is some anecdotal evidence to support
the possible use of cannabis or cannabinoids in the treatment of epilepsy, but little
more. Cannabinoids can exert both convulsant and anticonvulsant effects in various animal
tests. Of greatest interest are the anticonvulsant properties of the naturally occurring
cannabinoid cannabidiol; this compound is essentially devoid of the psychoactive effects
of THC. The limited clinical data available on the use of cannabidiol in the treatment of
epilepsy are, however, equivocal and based on very small numbers of patients. The BMA
report concludes, "It could possibly provide a useful adjunctive therapy for patients
poorly controlled on presently available drugs. THC and other psychoactive cannabinoids
are probably not suitable as anticonvulsants".
5.32 Cannabinoids cause a lowering of pressure in
the eye both in animals and in man, although the site of action and the mechanism involved
remain unknown. It has been suggested that cannabis or cannabinoids might be useful in the
treatment of elevated intraocular pressure (IOP) in glaucoma, one of the commonest causes
of blindness (see the BMA and WHO reports). Keith Green, Professor of Ophthalmology at the
Medical College of Georgia, USA, told us the results of his own studies in more than 300
human subjects with both normal and raised IOP. Cannabis caused an average 25 per
cent decrease in IOP which lasted for 3-4 hours. However, in order to maintain IOP at
baseline levels, patients would have to smoke as many as 10 cannabis cigarettes a day,
which is not practicable in view of the psychoactive effects of the drug and its ability
to impair cognitive function. Professor Green calls for further research to determine the
mechanisms involved, in order to see whether the desired ocular effects could be
dissociated from the intoxicant effects (p 219; cp Appendix 3, paragraph 12). A
similar view is expressed in the BMA and AMA reports, and by Professor Hall (Q 753).
5.33 Cannabis and THC dilate the small airways of
the lung, and this has suggested a possible application in the treatment of bronchial
asthma. However, according to the BMA report, there have been few clinical trials and
these were mostly in the 1970s before the advent of the more powerful drugs now available
for the treatment of this illness. Smoked cannabis is clearly unsuitable for the treatment
of asthma because of the irritant effects of the smoke, and THC delivered by aerosol also
appears to have irritating effects. The Royal Society, however, conclude,
"Cannabinoids...seem to be no less effective than conventional drug treatments.
Further studies are required to improve cannabinoid formulation for administration as an
aerosol" (p 294, cp Hall Q 753).
5.34 It is interesting to note that, if cannabis
were effective in both glaucoma and bronchial asthma, it would be especially useful for
patients suffering from both conditions, since many treatments for one of these conditions
are contra-indicated for the other.
Need for clinical trials of cannabis and cannabinoids
5.35 As noted above, the Government consider that
the burden rests on the proponents of wider medical use to satisfy the Medicines Control
Agency that the proposed medicine fulfils the normal criteria of quality, safety and
efficacy. Dr Brian Davis of the MCA (QQ 167-171) emphasised that efficacy can be
established only by undertaking controlled scientific trials; anecdotal evidence is not
acceptable. The BMA report and the Multiple Sclerosis Society (Q 389) accept this
5.36 The requirements for approval of a new
medicine are summarised as follows by the Royal Pharmaceutical Society (p 290):
" The active compound must be
characterised chemically and physically;
The active compound must be
presented in a standardised dosage formulation;
Adequate tests must have been
conducted on its safety;
Adequate controlled clinical
studies must have been conducted in well-defined disease entities and efficacy
The evidence must have been
published and subjected to peer review."
No-one claims that cannabis, or any cannabis-based medicine other than nabilone and
dronabinol, has yet passed any of these tests.
5.37 There have been few adequately controlled
clinical trials to date on cannabis and the cannabinoids, except as anti-emetics (see
above); those which have been published are listed in Appendix III to the BMA report. (For
details of what constitutes a clinical trial, see Box 7.) In MS, there have been only six
trials, with a total of only 41 patients. There is broad agreement that more and better
clinical trials would be a good thing (eg DH Q 180, ACT p 28). As Dr Pertwee
pointed out, there is an element of urgency: "Cannabis is already being used...We are
not in a situation where we can wait and see" (Q 317). The situation is
particularly urgent with respect to symptom control in MS, as the BMA report acknowledges,
because of a current lack of treatments. Similarly, in analgesia (pain control), there has
been no new drug for 20 years (Notcutt Q 411).
BOX 7: CONTROLLED CLINICAL TRIALS
|The approval of new medicines for human use requires that they be tested
rigorously in controlled clinical trials. "Controlled" means comparing the test
drug with an inactive dummy or "placebo". Placebo tablets or capsules are
prepared in such a manner that they cannot be distinguished from the active test drug. In
a "double-blind" placebo-controlled trial neither the patient nor the doctor or
nurse knows whether active drug or placebo is given to any particular patient; this
information is held in coded form by a person not actively involved in the conduct of the
trial and is not made available until the trial has ended. Patients are randomly allocated
to placebo and test drug groups to avoid any possible bias in the selection of those who
are to receive the active drug. The outcome of the trial should involve objective
measurements wherever possible, using predetermined outcome measures or
"endpoints". The success or failure of the trial is measured by criteria
established in a written trial protocol before the start of the trial. The trial should
include a sufficiently large number of subjects to provide statistically significant
differences in outcome measures between the placebo and drugtreated groups.
|A variant on the use of separate groups of patients to receive placebo or
test drug is the socalled "crossover" design, in which the same patients
receive placebo and test drug at different stages during the trial and are crossed over
from one to the other after a "wash-out" period in a random order, so that the
trial remains double-blind.
|The conduct of any clinical trial involving patients must be approved by
the Medicines Control Agency, who issue a Clinical Trial Certificate (CTX) if the detailed
written protocol for the trial meets with their approval. In addition the conduct of a
clinical trial requires the prior approval of the local Ethics Committee at the site where
the study is to be conducted.
5.38 The BMA report called on the Clinical
Cannabinoid Group (an informal network of interested researchers convened by Dr Pertwee),
patient groups, pharmaceutical companies and the Department of Health to "work
together to encourage" trials. In their written evidence the BMA say, "the
accumulation of scientific evidence has been hampered by regulations restricting the use
of cannabinoids to one clinical indication" [anti-emesis] (p 11). Following
their report, the BMA met the Government's Chief Medical Officer in March 1998 "to
discuss likely further actions in moving forward clinical trials of cannabinoids for
therapeutic uses". At the meeting, it was agreed that an appropriate body to conduct
such trials was required and that it should be an independent or institutional research
5.39 Clinical trials are expensive, but the
research funding bodies have no objection of principle to funding work in this field. The
MRC report a shortage of high-quality research proposals in this area (Q 629); but
they would be "supportive" of funding well-conceived clinical trials in this
field, and would even be prepared to consider a grant application "out of turn"
(QQ 638, 769). In February 1998 the MRC had three grants, one to Dr Pertwee and
two to Dr Kendall; Dr Pertwee's has now finished, as has one of
Dr Kendall's (Q 621). The Wellcome Trust had made nine grants since 1990: five
project grants, including three to Dr Pertwee, and four research career re-entry
fellowship grants, all to colleagues of Dr Pertwee. The Multiple Sclerosis Society and the
BMA are also willing to help fund a trial (MSSoc p 89; Q 769). The Department of
Health do not normally fund trials, but might "facilitate" (Q 194). Besides
the Wellcome Trust, Dr Pertwee's research group in Aberdeen is also funded by the
USA's National Institute on Drug Abuse (part of the National Institutes of Health) and a
pharmaceutical company (Q 252).
5.40 Professor Edwards (Q 19) questions the
justification for carrying out expensive controlled trials with cannabis. He is concerned
about the possibility of diversion to misuse of the drug. As a preliminary, he favours a
series of smaller-scale clinical investigations in individual patients.
5.41 Several of our witnesses have commented on the
difficulties of conducting clinical trials with cannabis. How can a standardised product
be made available? What formulation is to be used? How can the dose be predicted for any
particular medical condition? How consistent and predictable would blood levels of THC be
(QQ 7, 8, 180, 781)? In addition, individual patients are likely to differ
considerably in the dose needed to control their symptomsas with the use of opiates
in the control of severe pain, where Professor Wall points out that a tenfold range of
doses is commonly observed (QQ 112-3).
5.42 Professor David GrahameSmith, Chairman of
the Advisory Council on the Misuse of Drugs (Q 8), raised the question of the
difficulty of carrying out a doubleblind placebo-controlled trial with a psychoactive
agent, as the drug could easily be distinguished from the inert placebo. Professor Lader
suggested that one solution to this might be to test cannabis by comparison with some
other psychoactive drug, rather than against an inactive placebo. Other possibilities are
to use doses of the active drug too small to have psychoactive effect; or to proceed with
an inactive placebo, and find out at the end of the trial how far patients could tell
whether they were receiving the active drug or the placebo (Q 779).
5.43 In addition to these practical problems,
clinical researchers face extra legal hurdles, and a generally negative climate of
opinion, because of the status of cannabis as a Schedule 1 controlled drug. We
consider this problem, and what might be done about it, in Chapter 7.
5.44 Clinical trials are now under active
consideration in several fora. First, Dr Geoffrey Guy, a pharmaceutical entrepreneur, has
recently set up GW Pharmaceuticals, to conduct licensed research and develop
cannabis-based medicines, in collaboration with HortaPharm BV of Holland (see Dr Guy's
evidence, and QQ 107, 135, 413-420, 447). The aim is to produce standardised
whole-plant extracts, rather than single chemicals, from plants bred for standard
cannabinoid content, with a non-smoking mode of administration offering the advantages of
smoking without the harm, and to proceed via clinical trials to an application for a
product licence. Dr Guy received licences to cultivate cannabis, and to possess and
supply it for research, from the Home Office in June 1998. He is now recruiting patients
for trials, with help from the ACT.
5.45 Dr Guy is confident that rigorous trials
can be mounted, and that contamination of the plant material can be avoided. He advocates
the use of plant-derived products, and cites the examples of gentamicin, papaveretum and
digitalis as approved plantderived products that contain complex mixtures of alkaloids.
He believes that, by using controlled growing conditions and cloned cannabis plants, it
will be possible to produce a herbal preparation of consistent composition with adequate
quality controls. This position is reinforced by data from the Dutch organisation
Maripharm (see Appendix 4 paragraph 14), who have been able to produce medical-grade
herbal cannabis selected to have a consistent content of THC
(10.7 ± 0.1 per cent) and a low content of other cannabinoids.
5.46 Secondly, following the meeting between the
BMA and the CMO (see above, paragraph 5.38), the Royal Pharmaceutical Society have set up
a "working party on therapeutic uses of cannabinoids", chaired by Professor
Sir William Asscher, a past Chairman of the Committee on Safety of Medicines. The
group includes representatives of the NHS R&D Directorate, the MRC and the Multiple
Sclerosis Society, and researchers including Dr Pertwee. Its objectives are "to
produce guidelines for pilot clinical trials for cannabinoids as proof of principle of
their effectiveness, and to assist those who wish to conduct such trials to successfully
complete them and publish the results". The group has drawn up protocols for two
trials: one for spasticity arising from MS, the other for post-operative pain. In each
case the trial will be longitudinal (i.e. not a cross-over trial), involving three groups
of patients: one will be given dronabinol, another an extract of cannabis containing the
same quantity of THC, and the third a placebo. Smoking has been ruled out; administration
will be by oral capsule. The lead clinician for the pain trial is Dr Anita Holdcroft,
whose previous single-patient trial of cannabis was noted above; the lead clinician for
the MS trial is Dr John Zajicek of Derriford Hospital, Plymouth. The protocols are to be
launched shortly, at which point applications will be made for funding (from
non-industrial sources) and for Home Office licences. (See the evidence of members of the
working party, QQ 768-811.)
5.47 In addition, Jo Barnes of Exeter
University is launching a pilot study of oral THC involving 30 MS patients, funded by the
university, intended to "provide data which can be used for a sample size calculation
for a full-scale study" (p 217). Dr Robson is planning pilot studies using
nabilone and dronabinol for detoxification from opiates and as an anxiolytic/hypnotic in
acute drug-related problems (p 118, Q 458). Professor Wall knows of three other
United Kingdom trials at an advanced planning stage, by Dr Pertwee, Dr Notcutt
(Q 448), and Dr Clare Fowler at the National Hospital for Neurology and
Neurosurgery in London.
5.48 These various initiatives are, or may become,
interrelated. The Asscher group trials, and others, may use GW Pharmaceuticals as the
source of supply of cannabis material; in that case they might be covered by an extension
of Dr Guy's Home Office licence. The Exeter group, and others, may bring their trials
within the Asscher group's protocols, so as to become part of a national study. In the
end, it is possible that all or most of these initiatives will come together into two
national trials, using the Asscher group's protocols and Dr Guy's licence and materials.
Should clinical trials be limited to cannabinoids?
5.49 Both Dr Guy and the Asscher group propose to
conduct trials involving extracts of herbal cannabis; but according to several of our
witnesses this may be a mistake. Professor Ashton and Professor Nathanson of the BMA
(Q 55), reflecting the position of the BMA report itself, argued strongly in favour
of trials of synthetic cannabinoids rather than herbal cannabis, because of the
difficulties of obtaining standardised preparations of the plant material. Both Dr Guy and
the Asscher group believe that they can solve the problem of standardised preparations.
5.50 The Association of Chief Police Officers argue
that clinical research should be confined to individual cannabinoids: ACPO believe that
cannabis is a harmful substance, the control of which must be continued (p 196). The
Christian Institute take a similar view, arguing inter alia that medical use might
serve as a front for legalisation (p 208). (We consider this argument below in
Chapter 7.) Mary Brett, Head of Health Education at Dr Challoner's Grammar School, writes, "All scientific evidence is
unequivocal in favour of maintaining prohibition of crude marijuana for both medical and
recreational use. However, purified cannabinoids may, after rigorous testing and clinical
trials in comparison with other and existing treatments, prove to be beneficial in certain
disorders ..." (p 206).
5.51 Others, however, favour research on herbal
preparations derived from cannabis. Professor Wall argues in favour of trials of
cannabis rather than pure cannabinoids. He criticises the BMA report for recommending that
trials be confined to synthetic cannabinoids (p 32); he considers that it would be
premature at this stage of our knowledge to assume that the only active substance in
cannabis is THC (Q 103). We have received anecdotal evidence that users who have
tried cannabis and nabilone and/or dronabinol prefer cannabis (LMMSG p 271; ACT
pp 28, 30; IDMU p 228).
5.52 The Royal Society (p 295) also conclude
that "Several components of cannabis might be required to reproduce the effects seen
with the whole drug". Others in favour of including cannabis itself in any programme
of trials include the Royal Pharmaceutical Society (p 284), Dr Kendall
(p 268), Dr Pertwee (QQ 266, 315), Dr Robson (Q 480),
Dr Stewart and Dr Schon. The Multiple Sclerosis Society (Q 352) point out
that cannabis is available, and is what existing medical users are using; the ACT observe
that including it in trials would permit existing users to regularise their position (by
enrolling for trials) without changing their medication (Q 149).
Professor George Radda, Chief Executive of the MRC, would not rule out extracts of
herbal cannabis; "but we must know the composition" (Q 645).
5.53 Some witnesses point out that the variable
chemical composition of herbal cannabis can be turned to medical advantage. The London
Medical Marijuana Support Group argue that differing strains of cannabis containing
different proportions of THC, cannabinol (CBN) and cannabidiol (CBD) might have different
medical effects: "The more CBN and CBD, the greater the intensity of body related
sensations; the less CBN and CBD and the more THC, the more mentally active the
stimulation will generally be. High CBN and CBD cannabis is more effective for the control
of symptoms which are generally felt as being body related, such as chronic pain"
(p 270). Neil Montgomery also maintains that cannabis resins of different geographic
origin elicit distinct patterns of psychoactive effect (Q 594). There is, however, no
scientific evidence available on these topics.
Should clinical trials include smoking?
5.54 Both Dr Guy and the Asscher group have ruled
out smoking for the purposes of their trials; and many of our witnesses would support them
(e.g. Notcutt p 104, Henry p 224, RPharmSoc p 284, Wall Q 103, Pertwee
QQ 266, 315, MSSoc Q 364, ACT Q 154). Smoking is felt to carry too great a
potential health risk: see Chapter 4. However, as noted above, there are anecdotal
reports that those who use cannabis for medical purposes favour smoked cannabis over
orally administered cannabinoids such as nabilone. The perceived advantages of smoked
cannabis may be due to the rapid absorption and flexibility of dosecontrol offered by
smoking as a route of administration: see Chapter 3.
5.55 Dr Robson suggested that there should be a
comparison in clinical trials between smoked cannabis and smoked THC (Q 480). The
Asscher group's proposal, to compare orally administered THC with an orally administered
cannabis product, will achieve the same result, namely a comparison of like with like.
5.56 There is considerable discussion of possible
improvements in the mode of administration of cannabis and synthetic cannabinoids (e.g.
QQ 60, 266-273). IDMU (p 235) described recent research in the United States on the
ability of various methods of smoking herbal cannabis to reduce tar intake relative to
THC. Surprisingly, the use of a water pipe, in which the cannabis smoke is passed through
water prior to being inhaled, and the use of a vaporiser, in which herbal cannabis is
heated but not burned, had relatively little effect in reducing the amount of tar inhaled.
Unfortunately the slow and unreliable absorption of herbal cannabis and synthetic
cannabinoids taken by mouth can lead to both under and overdosing. Other possibilities
include the development of inhalers (e.g. Guy QQ 713-4), sprays, rectal suppositories
(see Appendix 3, paragraph 3) and skin patches, and a sub-lingual method (taking a
tincture under the tongueLMMSG p 270). Research on such alternative delivery
systems is held to be a high priority by many witnesses.
5.57 Although there is general agreement that
smoked cannabis carries a potential risk for long-term users, the medical application of
smoked cannabis is not ruled out by all. The US National Institutes of Health report says,
" ...there might be some patient populations, e.g. cancer patients experiencing
nausea and vomiting during chemotherapy, for whom the inhalation route might offer
advantages over the currently available capsule formulation [of THC]". They conclude,
"In summary, the testing of smoked marijuana to evaluate its therapeutic effects is a
difficult, but not impossible, task". The American Medical Association report
recommends "that adequate and well controlled studies of smoked marijuana be
conducted in patients who have serious conditions for which preclinical, anecdotal or
controlled evidence suggests possible efficacy including AIDS wasting syndrome, severe
acute or delayed emesis induced by chemotherapy, multiple sclerosis, spinal cord injury,
dystonia [involuntary muscle movements, e.g. a tic], and neuropathic pain...". Among
our witnesses, those who would include smoking in trials include Dr Schon
(p 304), Dr Stewart (p 305) and Dr Robson (Q 480); and
Professor Radda of the MRC would be prepared to do so, provided that the trial
protocols were satisfactory (QQ 646, 654).
21 Holdcroft A et al. Pain relief with oral
cannabinoids in familial Mediterranean fever. Anaesthesia, 1997, 52, 483. Back
22 Mrs Brett has written widely on cannabis, and advises the